The modulation of electromagnetic radiation is widely used, for example, to propagate signals carried at the modulation frequency, to improve signal-to-noise with phase-sensitive detection at the modulation frequency, or to switch light on and off. Many approaches well known in the art have been employed to effect the modulation without use of moving mechanical parts. These include electrical control of the radiation-source intensity itself, use of electro-optic materials to adjust the polarization orientation relative to the transmission axis of a polarizing filter, and use of Bragg diffraction at acoustic frequencies. See K. Y. Lau, N. Bar-Chaim, I. Ury, Ch. Harder, and A. Yariv, “Direct amplitude modulation of short-cavity GaAs lasers up to X-band frequencies,” Appl. Phys. Lett, 43 (1) (1983); and Louay Eldada, “Optical communication components,” Review Of Scientific Instruments, Vol. 75, No. 3, pp. 575-593 (2004). Modulating optical power by dynamically changing the reflectance or absorbance of a material has generally not been employed because of the difficulty of producing significant amplitude or frequency modulation.
A modulator exposed to broad-band illumination (e.g., white light) may also serve as a tunable source of radiation if it transmits or reflects a narrow band of wavelengths around a tunable center wavelength. This can be accomplished with commercial liquid-crystal tunable filters, but such filters have limited operation speeds (ms) and do not function in the mid- or far-IR spectral range. Other specialized acousto-optic tunable filters may operate in the IR, but require light to pass through bulky crystals and require substantial driving power. See N. B. Singh, D. Kahler, D. J. Knuteson, M. Gottlieb, D. Suhre, A. Berghmans, B. Wagner, J. Hedrick, T. Karr, and J. J. Hawkins, “Operational characteristics of a long-wavelength IR multispectral imager based on an acousto-optic tunable filter,” Opt. Engineering 2008, 47 (1), 013201.
When acting as a tunable source, a modulator can be used in molecular sensing applications if its wavelength can be tuned to optical absorptions characteristic of the analyte (i.e., the chemical of interest), as in optical absorption spectroscopy and infrared absorption spectroscopy (IRAS). Fieldable or remote sensing IRAS spectrometers, of particular interest to the DoD, but also of general commercial interest, are hampered by the paucity and small tuning range of available sources as well as their size, weight and power requirements.
Another sensor approach is based on surface-enhanced infrared absorption (SEIRA), where the detection of vibrational “fingerprints” of molecules adsorbed on the antenna is enhanced by the strong local optical fields near a rough or nanostructured surface, which could be a tunable nanoantenna. See R. Bukasov and J. S. Shumaker-Parry, “Silver nanocrescents with infrared plasmonic properties as tunable substrates for surface enhanced infrared absorption spectroscopy,” Anal. Chem. 2009, 81, 4531-4535; R. F. Aroca, D. J. Ross, and C. Domingo, “Surface-enhanced infrared spectroscopy,” Appl. Spectrosc. 2004, 58, 324A-338A; and M. S. Anderson, “Enhanced Infrared Absorption with Dielectric Nanoparticles,” Appl. Phys. Lett. 2003, 83 (14), 2964-2966.
Recently, miniaturization has been pursued for optical functions through the use of highly confined optical modes, which have potential to improve the performance and reduce the size and power requirements of optical modulators and sensors. For example, in sensing applications based on nanoantennas (e.g., SEIRA-based sensors), extreme miniaturization of the antenna is desirable both to increase the relative response of an individual antenna, and to increase the surface area for enhanced response-per-unit-area of the sensor platform.
Presently, highly confined optical modes are realized with surface-plasmons in metal nanostructures and waveguides. However, because plasmonic systems rely on free charge-carriers moving in response to optical fields (e.g., optical conduction currents in a metal), they suffer inherently from large scattering and absorption losses in the charge-carrier ensemble.
An alternative lower-loss approach, central to our disclosure, uses the vibrational motion of charge bound to the positive and negative atomic or molecular ions comprising a polar-dielectric (e.g., SiC, GaN, etc.) lattice. Charge bound to the positive and negative ions comprising a polar-dielectric (e.g., SiC, GaN, etc.) lattice will undergo a vibrational motion when stimulated by an outside force. This vibrational motion of atoms in the lattice is known as a “phonon,” while the coherent oscillatory motion of the charge carriers (i.e., of the free electron gas) is known as a “plasmon.” Each of these oscillatory motions has an associated wavelength, where λplasmon is the wavelength corresponding to the characteristic plasmon frequency of a metal, and λTO and λLO are the wavelengths associated with the transverse and longitudinal optical phonon vibrational frequencies, respectively.
In certain material-dependent wavelength ranges near the so-called “Reststrahlen” band, these polar-lattice vibrations in the material interact with light to produce surface phonon polaritons (SPhPs), which cause the optical response of the material (referred to herein as a “SPhP material”) to be similar to that of a metal, albeit without the presence of free carriers and the associated electrical conductivity and optical losses (due to fast carrier scattering rates). For both metals and SPhP materials, the real part ∈1(λ) of the complex dielectric function—the physical parameter governing the optical response—assumes negative values for certain wavelengths λ, lending high reflectance to metals when λ≧λplasmon and to SPhP materials when λTO≧λ≧λLO.
Negative values of ∈1(λ) permit resonant antennas to be constructed that are much smaller than their resonant wavelength, through what is sometimes called a Frolich, or dipole, mode. Such plasmonic antennas are referred to as local surface-plasmon resonators (LSPRs); here we refer to the surface phonon-polariton analog as a local surface-phonon resonator, and such an antenna is an “LSPhP” resonant antenna. This class of antenna resonates at a wavelength λres when ∈1(λres) assumes a specific value that depends on the geometric shape of the antenna. A well-known geometry is the sphere, which resonates when∈1(λres)=−2∈a,  (1)where ∈a is the dielectric constant of the ambient material surrounding the sphere, with ∈a=1 for an LSPhP resonator in air. Notably, principles of electromagnetism do not impose any lower limit to the size of these antennas because their resonant wavelength is set only by geometric shape. This is a unique distinction with the more common half-wave antennas and their kin, the size of which bears a fixed relationship to the resonant wavelength.
However, plasmonic antennas made of metals exhibit high losses. As a result, their resonant wavelengths have a broad spectral width which makes them less effective for applications requiring a strong response over a narrow wavelength band, such as thermal or quantum emitters or wavelength filters. Another shortcoming of plasmonic antennas made of metal is that their resonance wavelength cannot be dynamically tuned because the conduction electron density of a metal, which determines ∈1(λ) and hence the resonant frequency, is large and difficult to modify via traditional means (e.g., by electrostatic gating or optical pumping).
Although a plasmonic antenna made of metal can be tuned by altering the dielectric function of the nearby environment through the introduction of charge carriers, the benefits of large tuning ranges and narrow resonances in the infrared are not likely to be obtained. See Y. C Jun and I. Brener, “Electrically tunable infrared metamaterials based on depletion-type semiconductor devices” J. Opt. 2012, 14, 114013. Similarly, alterations in the nearby dielectric environment can be induced through electro-optic means, but again, tuning ranges will be severely limited. In principle, plasmonic antennas could also be constructed of semiconductors in which free carriers are introduced dynamically through electrical injection or optical pumping, see U.S. Patent Application Publication No. 2012/0074323 by J. Gomez rivas, V. Giannini, A. Berrier, S. A. Maier, M. Matters-Kammerer, and L. Tripodi, “THz frequency range antenna” (Mar. 29, 2012), but the density required to produce a Frolich resonance in the wavelength range of interest is prohibitively large and deleteriously lossy, where the losses result in resonance bands that are broad compared to the tuning range, which limits modulation depth.